Biphasic mechanisms of injury and repair in stroke therapy: The role of oxidative stress in stroke is well accepted. However, the recent SAINT-2 clinical trial of a free radical spin trap failed. Here, we propose the hypothesis that oxidative stress plays a biphasic role in stroke. Reactive oxygen species (ROS) are damaging in the acute phase. But they may be involved in neuro- vascular remodeling during stroke recovery. Our pilot data suggest that (i) ROS are required for angiogenesis in brain endothelial cultures; (ii) ROS are required for endothelial production of neuroprotective BDNF; (iii) delayed treatment with ROS scavengers in a rat stroke model suppresses vascular remodeling in peri-infarct cortex; and (iv) delayed ROS scavenging interferes with stroke recovery and ends up exacerbating tissue infarction. To build on these initial data, we will pursue the following specific aims. In AIM 1, we examine the molecular mechanisms that link ROS to angiogenesis. We will use brain endothelial cell cultures to test the hypothesis that SDF-1 triggers an NADPH oxidase (Nox) mechanism that upregulates endogenous ROS. Increased ROS then activates the MAP kinase called JNK, which goes on to upregulate neurovascular mediators such as MMP-9, VEGF and BDNF. The integrated contribution of these signals will be tested in endothelial proliferation, migration, and tube formation assays. Besides resident brain endothelium, circulating endothelial progenitor cells (EPCs) also contribute to neurovascular recovery after stroke. So in AIM 2, we investigate how ROS promotes EPC function. We hypothesize that the same SDF-1, Nox and JNK pathways serve to regulate MMP-9, VEGF and BDNF signals that underlie EPC proliferation, maturation, colony formation and tube formation. Hence, blockade of ROS or Nox is predicted to perturb EPC function. Finally, in AIM 3, we ask whether the cell culture effects studied in the first two aims can be validated in vivo. We hypothesize that early and short-term free radical scavenging will be protective, whereas delayed or prolonged scavenging will end up interfering with remodeling and recovery in rat or mouse stroke models. Similar pro-angiogenic signals will be examined in peri-infarct cortex as well as circulating EPCs. Optical and functional MR imaging will be applied to assess peri-infarct dynamics in vivo. Within the first few hrs after stroke, most events primarily mediate injury. By a few weeks, it is mostly remodeling. But what happens in between? In this project, we ask when damaged brain makes the transition from injury into repair, and how ROS might mediate injury versus benefit during this transition phase. This application is relevant to PA-08-099 Mechanisms of functional recovery after stroke.